Efficient insertion of exogenous nucleic acid into the chromosomal and extra-chromosomal nucleic acid of cells is desired in the art of molecular biology to identify chromosomal regions involved in expressing or regulating expression of peptides and proteins or to insert a nucleotide sequence of interest such as a primer binding site into a polynucleotide. This same technology is also advantageously used in developing new therapeutic and pharmacologic agents.
One common method relies upon in vivo Tn5 mutagenesis to insert polynucleotides of interest into cellular DNA and to construct libraries of cells that contain inserted polynucleotides at random or quasi-random locations. Existing in vivo Tn5 mutagenesis methods require target cells to encode transposase, either natively or from an introduced expression construct. Accordingly, it can be necessary to construct a suitable expression system appropriate to each target cell type. This can be time consuming, and requires extensive knowledge of the requirements of each target cell type.
In many cases, the gene that encodes transposase is encoded by an active transposon, which can continue to transpose in a target cell after the initial desired mutagenesis step. Such undesired residual transposition is undesired in that it complicates the analysis of insertional mutant libraries.
Furthermore, many techniques for in vivo Tn5 mutagenesis rely upon a complex biological mechanism for introducing exogenous DNA into the target cells, such as bacteriophage lambda transducing phage or a conjugating plasmid. It would be desirable to avoid requiring such complex biological systems.
The nature of natural Mu DNA, its ends, and its role in transposition are known. The left end includes 3 att repeat sites, denoted L1, L2 and L3 (all parallel). At the left end, only L1 is involved in transpososome formation. The nucleotide sequence of L1 is 5'-TGTATTGATTCACTTGAAGTACGAAAAAAA-3' (SEQ ID NO:1). L2 and L3 are spaced significantly apart from L1. The right end includes att repeat sites R1, R2 and R3. R1 and R2 are located close to one another and are involved in complex formation. The nucleotide sequence of R1 is 5'-TGAAGCGGCGCACGAAAAACGCGAAAGCGT-3' (SEQ ID NO:2). The nucleotide sequence of R2is 5'-GAAAGCGTTTCACGATAAATGCGAAAACTT-3' (SEQ ID NO:3). R3 and L1 are inverted relative to R1 and R2. MuA transposase, encoded by phage Mu, can bind to all 6 att repeat sites, but the MuA tetramer in a transpososome footprints on only L1, R.sub.1 and R2.
The Mu transposition reaction is most efficient when the Mu DNA includes one left end (containing L1, L2 and L3) and one right end (containing R1, R2 and R3). If the transposon ends are precut, strand transfer is most efficient if two right ends (containing R1 and R2) are used. If the non-transferred strand has a few (1 to 16 are equally effective) extra bases at the 5' end, then the reaction is even more efficient.
A commercial system for inserting a selectable artificial Mu transposon into target DNA is available from Finnzymes Oy and is based upon the above-noted considerations. In the commercial Mu system, MuA transposase and the target DNA are mixed together ex vivo to form products that have a polynucleotide inserted into target DNA. The precut transposon ends of non-transferred strands are provided with four extra bases. Both transposon ends are right ends that include R1 and R2, with the final TT of R2 absent from the construct. The sequence of the right ends used in the commercial product, presented as SEQ ID NO:4, is:
gatcTGAAGCGGCGCACGAAAAACGCGAAAGCGTTTCACGATAAATGCGAAAAC 3'-ACTTCGCCGCGTGCTTTTTGCGCTTTGCCAAAGTGCTATTTACGCTTTTG-5'
The transposition products are subsequently delivered (e.g., by electroporation) into competent target cells. Cells that contain such extra-chromosomal transposition products are then selected using well-known methods. Transposition is completed outside of the target cell. In the commercial embodiment, the Mu transposon does not transpose into the DNA of the target cell.
Shoji-Tanaka, A., et al., B.B.R.C. 203:1756-1764 (1994) describe using purified retroviral integrase to mediate gene transfer into murine cells. Kuspa, A. and W. F. Loomis, P.N.A.S. U.S.A. 89:8803-8807 (1992) and others have described specifically integrating a plasmid linearized with a restriction enzyme into a genomic restriction site by electroporating enzyme-cut nucleic acid along with the cleaving enzyme into target cells.